Shruthi Bandhu

Communication barriers should never limit opportunities.

Shruthi Bandhu originated from the realization that millions of deaf and mute individuals struggle with everyday interactions because sign language remains inaccessible to most people.

We set out to build an intelligent system capable of breaking this barrier, translating gestures dynamically, and making public environments, workspaces, and schools more inclusive through the power of AI.

We realized early on that building in isolation would lead to failure. We visited schools and classrooms to observe and listen to the real struggles and routines of silent communication.

Teachers highlighted that gesture translation must happen in sub-second timelines. Students pointed out that portable webcams or mobile apps are far better than bulky wearable sensors. These insights defined our hardware-agnostic product strategy.

There were no suitable datasets aligned with our vision.

Rather than adapting existing resources, we decided to build our own from scratch.

As a team, we learned sign language ourselves to better understand gesture nuances and communication patterns. We then designed and recorded our own dataset, capturing over 1,000 videos representing diverse gestures, perspectives, and real-world scenarios.

Every sample went through careful review and organisation before becoming part of our training pipeline.

This process taught us that building impactful AI systems often begins long before model training—it starts with deeply understanding the people and contexts the technology is meant to serve.

The patent journey taught us that impactful innovation extends beyond technology. It also requires strategic thinking, documentation, and protecting ideas responsibly.

Filing a patent allowed us to codify our architectural breakthroughs in gesture vector mapping and ensure our intellectual property is secure before presenting to institutional funds.